- Email: [email protected]
Selective detection and estimation of C-reactive protein in serum using surface-functionalized gold nano-particles
Analytica Chimica Acta 662 (2010) 186–192
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Selective detection and estimation of C-reactive protein in serum using surface-functionalized gold nano-particles Vidya Raj, K. Sreenivasan ∗ Laboratory for Polymer Analysis, Biomedical Technology Wing, Sree Chitra Tirunal Institute for Medical Sciences & Technology, Poojapura, Trivandrum 695012, Kerala, India
a r t i c l e
i n f o
Article history: Received 17 September 2009 Received in revised form 31 December 2009 Accepted 5 January 2010 Available online 11 January 2010 Keywords: Gold nano-particles O-phosphorylethanolamine Serum C-reactive protein
a b s t r a c t A new method for the detection of C-reactive protein (CRP) in serum using functionalized gold nano-particles (GNP) is reported. The affinity towards CRP is imparted to GNP by tethering Ophosphorylethanolamine (PEA) onto their surface. GNP and modified GNP were characterized using TEM, particle size analysis, zeta potential measurements, absorption spectroscopy and FT-IR techniques. The event of binding of CRP onto the PEA-GNP is followed by visibly observable colour change. We observed a red shift as well as a decrease in absorption in the plasmon peak of the modified GNP with the concentration of CRP. When the concentration of CRP exceeded 450 ng mL−1 , particles were aggregated and the solution became turbid. The method exhibited a linear range for CRP from 50 to 450 ng mL−1 with a detection limit of 50 ng mL−1 . The colour change and the variation in absorption of the GNP were highly specific to CRP even in the presence of albumin. We estimated CRP in blood serum collected from patients and the results obtained compared well with the estimation using the technique of nephelometry based on the antibody–antigen interaction. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Early and accurate diagnosis is crucial in designing strategies for the successful treatments of life-threatening ailment such as cardiac diseases (CD). One of the most powerful approaches designed and practised in the early diagnosis of many of the diseases is by the detection of certain molecules, known as markers, which show their presence in the body at a very early stage on the onset of the diseases. C-reactive protein (CRP) is one of the well-studied molecules in association with CD [1–3]. Concentration of CRP has been reported to be elevated up to 1000-folds during many infectious states including myocardial infraction [3]. Recent studies are suggestive that CRP along with serum cholesterol is decisive factors in the initiation and progression of CD [4]. A growing body of evidence has suggested that considerable prospect exists in using CRP measurement as a predictor of possible cardio vascular risks [5–7]. A CRP level up to 3 mg L−1 is considered normal, whereas a higher level is considered as indication of abnormality [8,9]. Over theses years several methodologies have been reported for its estimation [10–13]. Among the currently used techniques, immunoassays enjoy wide acceptance due to their specificity and sensitivity [14,15]. These techniques, which use specific antigen–antibody interactions, have been regarded as extremely useful approaches for the routine assay of CRP. One of the seri-
∗ Corresponding author. Tel.: +91 471 2520248; fax: +91 471 2341814. E-mail addresses: [email protected], [email protected] (K. Sreenivasan). 0003-2670/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2010.01.007
ous concerns of these methodologies is associated with the often noted non-specific binding and non-stability of the conjugated bio macromolecules (e.g. antibodies) [16]. Recently several measures have been taken to take care of these factors [17]. Efforts to develop techniques using non-biological components, interestingly, are scanty. Merit and Winkelman have reported an electrochemical sensor for CRP based on crown ether [18]. Very recently we reported the use of thermo responsive polymers conjugated with O-phosphorylethanolamine (PEA) for the detection of CRP [19]. Optical sensing based on the plasmon resonance absorption exhibited by nano-particles has been used with a view to develop analytical tools in clinical diagnosis [20–25]. In particular, extensive studies have been reported on the use of gold nano-particles as sensing platforms that exploit the plasmon resonance detection method for bio-specific interaction analysis as well as biomolecular interaction assays [26,27]. A small change in the size, shape, local environment, surface nature and degree of aggregation of nanoparticles leads to tunable changes in their optoelectronic properties which in fact enabled their use in sensing and quantification of molecules of interest [28]. Design of bifunctional ligands that will interact with nanoparticles on one hand and with the target analytes on the other is the primary requirement to suit gold nano-particles in sensing applications. Upon analyte–nano-particle interaction, a change in the collective properties can be used as chemical and biochemical sensors. Recently Jana, Ying and co-workers reported gold nanoparticles decorated with aptamer and antibodies to detect proteins
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
187
[29]. Appropriately functionalized gold nano-particles provide stable and continuous optical signals that can be used for the effective detection of bio molecules such as DNA and proteins. Herein we report a new sensing approach for CRP based on gold nano-particles modified by attaching PEA, the specific ligand for CRP with an objective to avoid antibodies. The detection of CRP was determined by changes in the plasmon resonance peak. The prospectus of the method is assessed by measuring CRP in blood serum collected from the patients reported to the Cardiology Department and the data generated were compared with the values obtained for the same samples using immunoassay technique. This report seems to be the first one for the detection of CRP using gold nano-particles without employing antibodies. 2. Experimental 2.1. Chemical Tetrachloroauric acid (III) trihydrate (HAuCl4 ·3H2 O), polyoxyethylene (20), sorbitan monolaurate (Tween 20), 16-mercaptohexadecanoic acid (16-MHDA), sodium citrate, sodium phosphate monobasic and sodium phosphate tribasic, O-phosphorylethanolamine, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), human albumin and human C-reactive protein were obtained from Sigma–Aldrich Chemicals Inc., Bangalore, India. These chemicals were used as received. Other analytical/chromatographic grade solvents were purchased from Merck or Sisco Chemicals, Mumbai, India. 2.1.1. Buffers and solutions All buffer solutions were prepared to 10 mM concentrations at pH 7. Deionized Millipore water was used in the preparations of all buffer solutions. The exact pH value of the buffer solutions was obtained using pH meter from Eutech Instruments. Solutions of 0.5 mM 16-MHDA were prepared in ethanol. Tween 20 solutions were prepared in sodium phosphate buffer at pH 7. For the activation of –COOH groups and for all other measurements, a working pH of 4.5 was used. 2.2. Instrumentation UV–vis spectroscopy was performed on a Varian model Cary win Bio 100 using 1 cm Quartz cuvette. Spectra were collected within a range of 400–800 nm. The infrared spectrum of nano-particles was measured in the range of 400–4000 cm−1 using a Nicollet Inc. (Madison, USA) model impact 5700 FT-IR spectrometer with a horizontal ATR accessory containing diamond crystal. The number of scans was 50. Samples were centrifuged and resuspended in buffer. The samples were then drop cast and the solvent was evaporated under IR lamp. This method yielded the IR spectra of nano-particles. Transmission electron microscopy (TEM) pictures of gold particle were taken using a Hitachi H600 TEM instrument. A finely focussed illuminating probe scanned the specimen and a standard Everhart-Thornley detector was used to collect the electrons transmitted through the specimen. The specimen holder with rotational grid facility was used with the microscope. Particle size distribution of the nano-particles was determined using photon correlation spectroscopy (PCS) on a Zetasizer nano-ZS (Malvern Instruments, Malvern, UK). Zeta potential () values were obtained using Malvern Zetasizer nano-ZS with a He–Ne laser beam. All measurements were done at a wavelength of 633.8 nm. Zeta potential was measured by applying an electric field across the dispersion. Particle within the dispersion with a zeta potential will migrate towards the electrode of opposite charge with a velocity proportional to the magnitude of the zeta
Scheme 1. Various steps involved in the conjugation of PEA onto gold nanoparticles.
potential. The frequency shift or phase shift of an incident laser beam caused by these moving particles is measured as the particle mobility, and this mobility is converted to the zeta potential using Smoluchowski or Huckel theories. MININEPHTM (Binding Site Ltd., Birmingham, UK) was used for determining CRP concentration in serum using the principle of nephelometry. The light source is a diode laser that emits at 670 nm. The focussed light passes through a cuvette containing the reaction mixture, where antibody/antigen complexes cause light to be scattered. This scatter is proportional to the amount of antibody/antigen complexes that have formed, and is detected by a photodiode. For each assay, a scatter is taken at the beginning of the antibody/antigen reaction (blank), followed by a second scatter reading at a fixed time. The analyte concentration is calculated using the difference between these two readings. The measuring range is 3.5–112 mg L−1 (3.5–112 g mL−1 ). 2.3. Synthesis and functionalization of gold nano-particles Gold nano-particles (GNP) were prepared following the method of Turkevich et al. [30]. Briefly, 95 mL of an aqueous chloroauric acid solution containing 5 mg of Au was brought to boil, and 5 mL of 1% sodium citrate solution was added to this boiling solution. The solution first changed to bluish colour, then purplish and eventually to ruby red. The solution was further boiled for 30 min and left to cool to room temperature. This method yielded spherical particles with a diameter of 39 ± 3 nm, as determined from TEM images. A series of surface modifications were done to functionalize the surface of gold nano-particles as depicted in Scheme 1. 2.3.1. Chemisorption of alkanethiols The surface of gold nano-particles was modified in the presence of non-ionic surfactant [31]. Briefly, equal volume (1000 L) of gold nano-particles and Tween 20 (1.83 mg/mL, before mixing) in 10 mM sodium phosphate buffer (pH 7) were gently mixed and allowed to stand for a minimum of 30 min. Then, 1000 L of 0.5 mM 16-MHDA solution was added and the final mixture was kept undisturbed for 3 h for 16-MHDA to be chemisorbed onto the gold nano-particles. Unreacted excess thiols and Tween 20 were removed from alkanethiol modified gold nano-particles
188
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
by repeated centrifugation for 10 min at 3000 rpm followed by decantation of supernatants. The nano-particles were then resuspended in buffer. To prevent light-induced flocculation of the nano-particles and oxidation of alkanethiolates; all gold nanoparticles solutions were stored in the dark and refrigerated at 4 ◦ C [32]. 2.3.2. Coupling of O-phosphorylethanolamine on to gold nano-particles Carbodiimide chemistry is commonly used to form amide bonds between carboxylic acid and primary amines [33]. As a first step, EDC reacts with a carboxylic acid group and forms an amine reactive intermediate. In the next step, addition of a primary aminecontaining molecule in the reaction mixture leads to an amide linkage between amine and the activated carboxylic acid group. The activation is generally done at lower pH (pH ∼ 4.5). Since no aggregation of particles was observed at this pH, for all experiments a working pH of 4.5 was selected [34]. A 10−2 M stock solution of EDC was prepared in deionized water and stored below 0 ◦ C. The stock solution was used immediately to prevent extensive loss of activity. 10−4 M stock solution of O-phosphorylethanolamine (PEA) was prepared in deionized Millipore water and stored at 4 ◦ C. Both these solutions were equilibrated to room temperature before use. To 1 mL of MHDA modified-gold nano-particles, 100 L each of stock solution of PEA and EDC were added step wise under vigorous stirring. Stirring was continued for 20–30 s and then the nano-particles were stored undisturbed for 24 h. This step yielded nano-particles containing PEA, the specific ligand for CRP. Unreacted reagents were removed by repeated centrifugation and the collected nano-particles were resuspended in buffer as described above. 2.3.3. Interaction of CRP with functionalized gold nano-particles A stock solution of CRP was prepared by dissolving 0.18 mg CRP in 50 mL of 0.1 M calcium chloride solution. To 1 mL of functionalized gold nano-particles conjugated with PEA, 20 L of stock solution of CRP was added. All the measurements were made in triplicate and the average was taken. The relative optical density (OD) was corrected for dilution. The OD of the peak was measured by adding varying amounts of CRP. 2.3.4. Interaction of albumin with functionalized gold nano-particles Stock solution of albumin was prepared by dissolving 100 mg albumin in 100 mL of 10 mM phosphate buffer. To 1 mL of functionalized gold nano-particles, 10 L of stock solution of human albumin solution was added. To the same solution varied quantity of CRP was added to know the binding of CRP in the presence of albumin. In each case, the absorption spectra were recorded. 2.3.5. Estimation of CRP in blood serum using functionalized nano-particles Blood samples were collected from patients reported to Cardiology Department. The samples were collected according to the guidelines [35]. 1 mL of PEA conjugated gold nano-particles (PEAGNP) was taken and 10 L serum separated from the blood samples was added to the solution. The solution was shaken well for 10 s, and the plasmon peak was monitored. All the measurements were made in triplicate. 3. Results and discussion 3.1. Synthesis and surface modifications of gold nano-particles GNPs synthesised by the method of Turkevich et al. [30] yielded spherical particles with an average diameter of 39 ± 3 nm by the
Fig. 1. Optical absorption spectra of gold nano-particles and modified nanoparticles.
particle size analysis. Changes in the plasmon resonance peak monitored during each stage of modifications are shown in Fig. 1. The surface plasmon resonance peak of unmodified gold nano-particles is centered on 520 nm as reported earlier indicating that the nanoparticles were not aggregated but well dispersed as individual particles. GNPs modified with 16-MHDA exhibit a peak around 524 nm. The red shift in the position of the plasmon absorption is produced by a perturbation in the dielectric constant around the nano-particles due to chemisorption of 16-MHDA forming a selfassembled monolayer [31]. No significant broadening was observed after this step, indicating that the particles were not aggregated upon chemisorption of 16-MHDA. Further modification with EDC and PEA resulted in red shift to 526 nm. The results were further confirmed by TEM imaging, particle size analysis, potential measurement and FT-IR analysis. The generation of spherical nano-particles with nearly identical sizes was substantiated from the TEM image shown in Fig. 2A. Detectable variation in particle size was resulted by each modification steps as reflected in TEM images shown in Fig. 2B and C. Particle sizes and potential parameters summarized in Table 1 further support the results emerged from TEM. Particle sizes were steadily increased after each modification steps namely formation of 16MHDA layer followed by the coupling of PEA. Furthermore, the potential values (Table 1) altered after each steps indicating the modification of the particles. The net negative charge of the gold nano-particles reduced as result of the attachment of 16-MHDA. The size also increased further reflecting the presence of MHDA on the particle surface. Even after the modification, presence of net negative charge on the nano-particle surface stabilized them against aggregation in water. The net negative charge is slightly increased when PEA is attached onto the particles through the –COOH groups of 16-MHDA. In weakly acidic conditions the phosphate moiety in PEA exists in anionic form (PO4 − ). The potential value of O-phosphorylethanolamine at pH 4.5 was found to be −0.367 (Fig. 1S, Supplementary information). The increase in net negative charge of GNP modified with PEA could be assigned to the additional charge acquired from phosphate groups. To get further insight into the surface modifications of gold nano-particles, FT-IR spectroscopy was used. Nano-particles functionalized with 16-MDHA showed a strong band at 1734 cm−1 indicating the presence of carboxylic acid functionalities on the parTable 1 potential and hydrodynamic size of gold nano-particles and modified nanoparticles in solution as a function of surface modifications. System
potential (mV)
Effective diameter (nm)
As prepared gold nano-particles (citrate coated) 16-MHDA functionalized
−65 ± 2
39 ± 3
−17.6 ± 1
45 ± 1
Nano-particle PEA-GNP CRP–PEA-GNP
−23.5 ± 2 −4.65 ± 2
51 ± 1 107 ± 5
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
189
Fig. 2. TEM micrograph of gold nano-particles during each stages of modification. (A) Gold nano-particles; (B) after modification with 16-MHDA; (C) PEA-GNP.
ticles (Fig. 2S, Supplementary information). –CO-stretching band of 16-MDHA has nearly disappeared in the next step, namely the conjugation of PEA through the activated –COOH groups. The coupling was further confirmed by the peak centered on 1572 cm−1 reflecting the formation of –OCNH entity (Supplementary information).
3.2. Binding of CRP onto the modified nano-particles We studied the interaction of CRP with the functionalized nano-particles. The PEA-GNPs were equilibrated with solutions containing varied concentrations of CRP. The binding of CRP
Fig. 3. Colour changes of the nano-particles induced by surface modification. (A) Gold nano-particles, (B) gold nano-particles modified with 16-MHDA, (C) gold nano-particles attached to PEA and (D) after addition of CRP.
190
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
Fig. 4. Optical absorption spectra of PEA-GNP in the presence of varied concentrations of CRP.
resulted in a visibly observable colour change, reflecting shift in the plasmon resonance absorption (Fig. 3). The optical absorption spectra of the nano-particles in the presence of varied amounts of CRP are shown in Fig. 4. A red shift of the plasmon peak with concomitant broadening of the spectrum occurred in a concentration-dependent manner. The shift and broadening attained maximum values at intermediate concentrations of CRP. CRP causes a change in the absorbance since it induces aggregation of gold nano-particles through specific molecular interactions with PEA. The degree of aggregation is concentration dependent with maximum aggregation occurring for intermediate concentration of CRP in the system. If only a small number of CRP molecules are present in the system, aggregation will not proceed to a significant extent because of the limited availability of the molecule that bridges nano-particle together. Only at intermediate concentration of CRP, when one CRP molecule can bind several gold nano-particles aggregation occurs since a single CRP molecule has five binding sites. When the concentration of CRP is large enough, PEA groups become unavailable for CRPmediated aggregation. Beyond a certain amount of CRP, the solution turned turbid indicating clumping of particles. Earlier studies have shown similar results, namely the aggregation of nano-particles mediated by proteins [36]. We used particle size measurement and TEM imaging to get further insight in to the process. The effective diameter of the protein bound particles increased to 107 ± 5 and zeta potential is reduced to −4.65 ± 2 (see Table 1). The reduced charge together with the increased size facilitates interaction of the particles resulting in the aggregation. TEM micrographs of PEAGNP after the addition of lower and higher concentrations of CRP are shown in Fig. 5A and B. The number of particles in the cluster increases significantly, when PEA-GNP was allowed to interact with CRP. The formation of bigger clusters was a manifestation of specific molecular recognition occurring between the multivalent CRP and PEA group in PEA-GNP. CRP molecules can link several PEA-GNP nano-particles together to form larger clusters. The aggregation with increase in CRP concentration was reasoned due to the bridging of nano-particles by the protein. We have drawn such a conclusion by visualizing the particles depicted in TEM (Fig. 5B). We recorded the infrared spectrum of the particles after equilibrating with CRP. The IR spectrum showed additional peaks at 1650 cm−1 and 1540 cm−1 , characteristic of amide I and amide II bands in protein chains, strongly suggesting the adsorption of CRP onto the PEA-GNP (Fig. 3S, Supplementary information). 3.3. Quantification of CRP The optical density measurements of PEA-GNPs at different CRP concentrations were monitored. It was found that with increase in CRP concentration the plasmon resonance peak showed red shift with a concomitant decrease in OD. Red shift in absorption peak and a reduction in intensity with concentration of analytes are well documented [37]. The change in OD values was increased initially and
Fig. 5. TEM micrograph of gold nano-particles after interacting with CRP. (A) On addition of lower concentration of CRP; (B) on addition of higher concentrations of CRP.
then levelled off indicating the saturation of available binding sites for CRP (Fig. 4S, Supplementary information). A linear relationship between OD of the plasmon peak and concentration was observed until CRP was bound onto the entire available PEA. Further addition of CRP did not show any variation in OD (Fig. 4S, Supplementary information). We observed that GNPs were started aggregating (slight turbidity) when the concentration of CRP exceeded 500 ng mL−1 and the solution became completely turbid beyond 600 ng mL−1 of CRP. Larger concentrations of CRP (>600 ng mL−1 ) caused the aggregation of almost all PEA-GNPs indicating the saturation of available binding sites for CRP. The change in absorption reaches maximum with the addition of 450–500 ng mL−1 of CRP. This observation led us to conclude that, this methodology can be used for the estimation of CRP in solutions having a concentration of ≤450 ng mL−1 . It was observed that PEA-GNPs are separated from each other before the addition of CRP, and they become closer and closer with the addition of increased amount of CRP leading to the clustering of particles. For quantification of CRP, we constructed a calibration plot using the concentrations and corresponding OD (Fig. 5S, Supplementary information).
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
191
Table 2 Quantitative data of CRP in serum—comparison of the methods. Sample code
Amount of CRP (nephelometry)a (mg L−1 )
Amount of CRP using (PEA-GNP) (mg L−1 )
A B C D E F G
39.07 22.72 16.79 12.76 11.45 Negative Sample Negative Sample
38.97 ± 0.21 22.45 ± 0.10 16.05 ± 0.50 12.15 ± 0.09 11.32 ± 0.12 No measurable shift in the peak No measurable shift in the peak
a
Immunoassay (average of three measurements).
3.4. Interaction of albumin with the modified particles Albumin is one of the major blood proteins. Non-specific adsorption of albumin onto the particles is known to be a major problem associated with the usage of functionalized gold nanoparticles. Albumin concentration is several folds higher than the concentration of CRP in the blood. Interference of albumin in the measurement, if any, has to be assessed. To the solution of PEA-GNPs, 10 L of stock solution of albumin was added and the plasmon resonance absorption was monitored. The plasmon absorption peak of PEA-GNP was centered at 526 nm. When albumin was added to the above solution, the plasmon peak is again centered at 526 nm indicating that albumin was not binding on to the ligand (Fig. 6S, Supplementary information). Park et al. have recently shown that binding of albumin onto polymeric nano-particles could be reduced remarkably by modifying the surface by phosphorylcholine moieties [17]. Also the isoelectric point of albumin is ∼4–4.5. At a pH below their isoelectric point, proteins carry a net positive charge; above their isoelectric point they carry a net negative charge. Since the working pH was 4.5 albumin is supposed to be negatively charged and hence adsorption of albumin onto the negatively charged PEA-GNP is unlikely. Interestingly, when CRP was added to the same solution, plasmon resonance peak shifted to 529 nm strongly suggesting the binding of CRP to its ligand even though albumin was present (Fig. 7S, Supplementary information). This observation further confirm that albumin is neither affecting the plasmon absorption peak nor interfering the interaction of CRP with its ligand. The plasmon absorption peak showed red shifts with increase in the concentration of CRP when additional amount of CRP was added into the same solution (Supplementary information). 3.5. Measurement of CRP in real samples Blood samples collected from various patients were tested using PEA-GNP and the results were compared with those obtained by normal CRP diagnostic test using CRP specific antibody. The clinical assay for CRP detection is generally carried out by immunoassay using nephelometry with detection limit of ∼3.5 g mL−1 . Serum samples containing <3.5 g mL−1 of CRP, as per this method, is considered as negative. Higher concentration of CRP (>3.5 g mL−1 ) is treated as higher risk factor for myocardial infraction and those samples are designated as positive samples. Here we provide a comparative study of CRP estimation using nephelometry method and PEA-GNPs. PEA-GNPs showed plasmon absorption peak around 526 nm (Fig. 6A). The samples designated as negative CRP samples did not show any considerable shift in the plasmon peak indicating very low concentration of CRP (Fig. 6A, Sample F in Table 2, see the overlapped traces). The plasmon peak, however, shifted to higher wavelength when serum termed as positive is added into the functionalized nano-particles reflecting the sensing of CRP by the particles. The plasmon peak was shifted to 529 nm for CRP concentration in the range of 11–12 g mL−1 (spectra D and E in Fig. 6B and Samples D and E in Table 2). The absorption
spectra of these two samples are almost overlapped since the CRP content is very close. With serum containing higher concentrations of CRP the spectra showed further shift with a decrease in intensity (spectra A, B and C, Fig. 6B and Samples A, B and C in Table 2). These samples were diluted several times in order to bring them in the sensing range. Table 2 compares quantitative data generated using both methods. The results emerged from the current method compare well with immunoassay method, indicating the reliability of the newly proposed approach using functionalized nano-particles. One of the remarkable features of the present method is its sensitivity comparing to the method based on nephelometry. The present approach enables the measurement in the nanogram range while the immunoassay is in the microgram range. The volume of serum used in the measurement was 10 L. Many serum samples containing higher quantity of CRP (as per immunoassay) were diluted prior to the analysis using the current approach since higher quantity of CRP led to the aggregation of PEA-GNPs. Several recent studies confirm that elevated serum level of plasma CRP is associated with an increased risk of experiencing myocardial infraction and sudden cardiac death in apparently healthy subjects. These studies undoubtedly suggest that elevated CRP concentration may predict a higher risk for future cardiovascular diseases [1–3]. Very recently Zhan and Bard reported a chemiluminescence method for the determination of CRP using polyclonal human CRP antibodies incorporated in liposomes [38]. These approaches though sensitive require antibodies as recognition element for CRP. The effort in this communication was to develop a methodology avoiding antibodies. Towards this goal we synthesised gold nanoparticles and functionalized the surface of gold nano-particles with CRP specific ligand. The data emerged from the study suggest that appropriately functionalized gold nano-particles can be employed
Fig. 6. (A) Optical absorption spectra of PEA-GNP before and after the addition of 10 L negative serum. (B) Optical absorption spectra of PEA-GNP (traces A, B, C, D and E) after adding 10 L serum samples.
192
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
for the assay of CRP in serum. A visibly observable colour change with positive samples is another added advantage of the present method in the sense that it could be used for the quick screening of positive and negative samples. 4. Conclusion We have developed a new colorimetric sensing approach employing functionalized gold nano-particles for the estimation of CRP in serum with a high degree of selectivity and sensitivity. Gold nano-particles were synthesised and functionalized with PEA, the specific ligand of CRP. The nano-particles were characterized using TEM, particle size, potential measurements, FT-IR and optical absorption spectroscopy. The plasmon absorption spectra of PEA-GNPs were found to show a considerable red shift in the presence of CRP reflecting its binding onto the particles. potential and particle sizes were found to vary in accordance with the modification and CRP binding. The concentration depended shift facilitated the measurement of CRP in serum. The binding of CRP onto the nano-particles was clearly visible with a distinct colour change. The newly designed approach was used to estimate the CRP in serum of patients and the values were compared well with those generated using immunoassay technique. The method involves simple synthetic approaches using commercially available components and is devoid of antibodies. Acknowledgements We wish to thank Dr. C.P. Sharma and Willy Paul for zeta potential and particle size measurement, Dr. Annie John for TEM analysis and Ms. Moly Antony for providing blood samples. Vidya Raj thanks Council of Scientific and Industrial Research (CSIR), India for the senior research fellowship.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
[21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]
Appendix A. Supplementary data
[33] [34] [35]
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.aca.2010.01.007.
[36] [37] [38]
P.M. Ridker, J.E. Buring, N.R. Cook, N. Rifai, Circulation 107 (2003) 391. W.T. Tillet, W.T.J. Francis, J. Exp. Med. 52 (1930) 561–567. I. Kushner, M.L. Broder, D.J. Karp, J. Clin. Invest. 61 (1978) 232–235. P.M. Ridker, Circulation 107 (2003) 363–369. Q. Zen, W. Zhong, R.F. Mortensen, J. Cell. Biochem. 64 (1997) 140–151. J.E. Volanaski, Mol. Immunol. 38 (2001) 189–197. J. Whicher, L. Biasucci, N. Rifai, Clin. Chem. Lab. Med. 37 (1999) 495–503. O. Gul, E. Calay, U. Sezerman, H. Basaga, Y. Gurbuz, Sens. Actuators B 126 (2007) 581–587. M.H.F. Meyer, M. Hartman, Biosens. Bioelectron. 21 (2006) 1987–1992. C. Glad, A.O. Grubb, Anal. Biochem. 85 (1978) 180–187. K. Kriz, F. Ibraimi, M. Lu, L. Hansson, D. Kriz, Anal. Chem. 77 (2005) 5920–5924. W. Miao, A.J. Bard, Anal. Chem. 75 (2003) 5825–5834. H.Y. Tsai, C.F. Hsu, I.W. Chiu, C. Bor Fuh, Anal. Chem. 79 (2007) 8416–8419. R.E. Banks, Crit. Rev. Clin. Lab. Sci. 37 (2000) 131–182. J.L. Bock, Am. J. Clin. Pathol. 113 (2000) 628–646. P.M. Ridker, M. Cushman, M.J. Stampfer, R.P. Tracy, C.H. Hennekens, Circulation 97 (1998) 425–428. A. Park, S. Kurosawa, J. Watanabe, K. Ishihara, Anal. Chem. 76 (2004) 2649–2655. C.M. Merritt, J.W. Winkelman, Anal. Chem. 61 (1989) 2362–2365. V. Raj, P.R. Hari, K. Sreenivasan, Anal. Chim. Acta 592 (2007) 45–50. V.H. Perez-Luna, K. Aslan, P. Betala, Colloidal gold, in: J.A. Schwarz, C.I. Contescu, K. Putyera (Eds.), Encyclopedia of Nanoscience and Nanotechnology, vol. 2, Marcel Dekker, New York, USA, 2004, pp. 27–49. U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters. Springer Series in Material Science 25, Springer, New York, USA, 1995. M.A. Hayat (Ed.), Colloidal Gold: Principles, Methods, and Applications, Academic Press, San Diego, CA, USA, 1989. M.C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293–346. E. Katz, I. Willner, Angew. Chem. Int. Ed. 43 (2004) 6042–6108. W. Knoll, Annu. Rev. Phys. Chem. 49 (1998) 569–638. N. Nath, A. Chilkoti, Anal. Chem. 74 (2002) 504–509. N. Nath, A. Chilkoti, Anal. Chem. 76 (2004) 5370–5378. J.J. Li, Y.A. Wang, W. Guo, J.C. Keay, T.D. Mishima, M.B. Johnson, X. Peng, J. Am. Chem. Soc. 125 (2003) 12567–12611. E. Nandan, N.R. Jana, J.Y. Ying, Adv. Mater. 20 (2008) 2068–2073. J. Turkevich, P.C. Stevenson, J. Hillier, Discuss. Faraday Soc. 11 (1981) 55–75. K. Asian, V.H. Perez-Luna, Langmuir 18 (2002) 6059–6065. C.S. Weisbecker, M.V. Merritt, G.M. Whitesides, Langmuir 12 (1996) 3763–3772. J. Lee, A.O. Govorov, J. Dulka, N.A. Kotov, Nano Lett. 4 (2004) 2323–2327. A. Gole, C.J. Murphy, Langmuir 21 (2005) 10756–10762. S.M. Lewis, A. Koepke, Haematology Laboratory Management and Practice, American Chemical Society, Washington, DC, 1995. K. Asian, C.C. Luhrs, V.H. Perez-Luna, J. Phys. Chem. B 108 (2004) 15631–15639. K. Ai, Y. Liu, L. Lu, J. Am. Chem. Soc. 131 (2009) 9496–9497. W. Zhan, A.J. Bard, Anal. Chem. 79 (2007) 459–463.
Contents lists available at ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Selective detection and estimation of C-reactive protein in serum using surface-functionalized gold nano-particles Vidya Raj, K. Sreenivasan ∗ Laboratory for Polymer Analysis, Biomedical Technology Wing, Sree Chitra Tirunal Institute for Medical Sciences & Technology, Poojapura, Trivandrum 695012, Kerala, India
a r t i c l e
i n f o
Article history: Received 17 September 2009 Received in revised form 31 December 2009 Accepted 5 January 2010 Available online 11 January 2010 Keywords: Gold nano-particles O-phosphorylethanolamine Serum C-reactive protein
a b s t r a c t A new method for the detection of C-reactive protein (CRP) in serum using functionalized gold nano-particles (GNP) is reported. The affinity towards CRP is imparted to GNP by tethering Ophosphorylethanolamine (PEA) onto their surface. GNP and modified GNP were characterized using TEM, particle size analysis, zeta potential measurements, absorption spectroscopy and FT-IR techniques. The event of binding of CRP onto the PEA-GNP is followed by visibly observable colour change. We observed a red shift as well as a decrease in absorption in the plasmon peak of the modified GNP with the concentration of CRP. When the concentration of CRP exceeded 450 ng mL−1 , particles were aggregated and the solution became turbid. The method exhibited a linear range for CRP from 50 to 450 ng mL−1 with a detection limit of 50 ng mL−1 . The colour change and the variation in absorption of the GNP were highly specific to CRP even in the presence of albumin. We estimated CRP in blood serum collected from patients and the results obtained compared well with the estimation using the technique of nephelometry based on the antibody–antigen interaction. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Early and accurate diagnosis is crucial in designing strategies for the successful treatments of life-threatening ailment such as cardiac diseases (CD). One of the most powerful approaches designed and practised in the early diagnosis of many of the diseases is by the detection of certain molecules, known as markers, which show their presence in the body at a very early stage on the onset of the diseases. C-reactive protein (CRP) is one of the well-studied molecules in association with CD [1–3]. Concentration of CRP has been reported to be elevated up to 1000-folds during many infectious states including myocardial infraction [3]. Recent studies are suggestive that CRP along with serum cholesterol is decisive factors in the initiation and progression of CD [4]. A growing body of evidence has suggested that considerable prospect exists in using CRP measurement as a predictor of possible cardio vascular risks [5–7]. A CRP level up to 3 mg L−1 is considered normal, whereas a higher level is considered as indication of abnormality [8,9]. Over theses years several methodologies have been reported for its estimation [10–13]. Among the currently used techniques, immunoassays enjoy wide acceptance due to their specificity and sensitivity [14,15]. These techniques, which use specific antigen–antibody interactions, have been regarded as extremely useful approaches for the routine assay of CRP. One of the seri-
∗ Corresponding author. Tel.: +91 471 2520248; fax: +91 471 2341814. E-mail addresses: [email protected], [email protected] (K. Sreenivasan). 0003-2670/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2010.01.007
ous concerns of these methodologies is associated with the often noted non-specific binding and non-stability of the conjugated bio macromolecules (e.g. antibodies) [16]. Recently several measures have been taken to take care of these factors [17]. Efforts to develop techniques using non-biological components, interestingly, are scanty. Merit and Winkelman have reported an electrochemical sensor for CRP based on crown ether [18]. Very recently we reported the use of thermo responsive polymers conjugated with O-phosphorylethanolamine (PEA) for the detection of CRP [19]. Optical sensing based on the plasmon resonance absorption exhibited by nano-particles has been used with a view to develop analytical tools in clinical diagnosis [20–25]. In particular, extensive studies have been reported on the use of gold nano-particles as sensing platforms that exploit the plasmon resonance detection method for bio-specific interaction analysis as well as biomolecular interaction assays [26,27]. A small change in the size, shape, local environment, surface nature and degree of aggregation of nanoparticles leads to tunable changes in their optoelectronic properties which in fact enabled their use in sensing and quantification of molecules of interest [28]. Design of bifunctional ligands that will interact with nanoparticles on one hand and with the target analytes on the other is the primary requirement to suit gold nano-particles in sensing applications. Upon analyte–nano-particle interaction, a change in the collective properties can be used as chemical and biochemical sensors. Recently Jana, Ying and co-workers reported gold nanoparticles decorated with aptamer and antibodies to detect proteins
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
187
[29]. Appropriately functionalized gold nano-particles provide stable and continuous optical signals that can be used for the effective detection of bio molecules such as DNA and proteins. Herein we report a new sensing approach for CRP based on gold nano-particles modified by attaching PEA, the specific ligand for CRP with an objective to avoid antibodies. The detection of CRP was determined by changes in the plasmon resonance peak. The prospectus of the method is assessed by measuring CRP in blood serum collected from the patients reported to the Cardiology Department and the data generated were compared with the values obtained for the same samples using immunoassay technique. This report seems to be the first one for the detection of CRP using gold nano-particles without employing antibodies. 2. Experimental 2.1. Chemical Tetrachloroauric acid (III) trihydrate (HAuCl4 ·3H2 O), polyoxyethylene (20), sorbitan monolaurate (Tween 20), 16-mercaptohexadecanoic acid (16-MHDA), sodium citrate, sodium phosphate monobasic and sodium phosphate tribasic, O-phosphorylethanolamine, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), human albumin and human C-reactive protein were obtained from Sigma–Aldrich Chemicals Inc., Bangalore, India. These chemicals were used as received. Other analytical/chromatographic grade solvents were purchased from Merck or Sisco Chemicals, Mumbai, India. 2.1.1. Buffers and solutions All buffer solutions were prepared to 10 mM concentrations at pH 7. Deionized Millipore water was used in the preparations of all buffer solutions. The exact pH value of the buffer solutions was obtained using pH meter from Eutech Instruments. Solutions of 0.5 mM 16-MHDA were prepared in ethanol. Tween 20 solutions were prepared in sodium phosphate buffer at pH 7. For the activation of –COOH groups and for all other measurements, a working pH of 4.5 was used. 2.2. Instrumentation UV–vis spectroscopy was performed on a Varian model Cary win Bio 100 using 1 cm Quartz cuvette. Spectra were collected within a range of 400–800 nm. The infrared spectrum of nano-particles was measured in the range of 400–4000 cm−1 using a Nicollet Inc. (Madison, USA) model impact 5700 FT-IR spectrometer with a horizontal ATR accessory containing diamond crystal. The number of scans was 50. Samples were centrifuged and resuspended in buffer. The samples were then drop cast and the solvent was evaporated under IR lamp. This method yielded the IR spectra of nano-particles. Transmission electron microscopy (TEM) pictures of gold particle were taken using a Hitachi H600 TEM instrument. A finely focussed illuminating probe scanned the specimen and a standard Everhart-Thornley detector was used to collect the electrons transmitted through the specimen. The specimen holder with rotational grid facility was used with the microscope. Particle size distribution of the nano-particles was determined using photon correlation spectroscopy (PCS) on a Zetasizer nano-ZS (Malvern Instruments, Malvern, UK). Zeta potential () values were obtained using Malvern Zetasizer nano-ZS with a He–Ne laser beam. All measurements were done at a wavelength of 633.8 nm. Zeta potential was measured by applying an electric field across the dispersion. Particle within the dispersion with a zeta potential will migrate towards the electrode of opposite charge with a velocity proportional to the magnitude of the zeta
Scheme 1. Various steps involved in the conjugation of PEA onto gold nanoparticles.
potential. The frequency shift or phase shift of an incident laser beam caused by these moving particles is measured as the particle mobility, and this mobility is converted to the zeta potential using Smoluchowski or Huckel theories. MININEPHTM (Binding Site Ltd., Birmingham, UK) was used for determining CRP concentration in serum using the principle of nephelometry. The light source is a diode laser that emits at 670 nm. The focussed light passes through a cuvette containing the reaction mixture, where antibody/antigen complexes cause light to be scattered. This scatter is proportional to the amount of antibody/antigen complexes that have formed, and is detected by a photodiode. For each assay, a scatter is taken at the beginning of the antibody/antigen reaction (blank), followed by a second scatter reading at a fixed time. The analyte concentration is calculated using the difference between these two readings. The measuring range is 3.5–112 mg L−1 (3.5–112 g mL−1 ). 2.3. Synthesis and functionalization of gold nano-particles Gold nano-particles (GNP) were prepared following the method of Turkevich et al. [30]. Briefly, 95 mL of an aqueous chloroauric acid solution containing 5 mg of Au was brought to boil, and 5 mL of 1% sodium citrate solution was added to this boiling solution. The solution first changed to bluish colour, then purplish and eventually to ruby red. The solution was further boiled for 30 min and left to cool to room temperature. This method yielded spherical particles with a diameter of 39 ± 3 nm, as determined from TEM images. A series of surface modifications were done to functionalize the surface of gold nano-particles as depicted in Scheme 1. 2.3.1. Chemisorption of alkanethiols The surface of gold nano-particles was modified in the presence of non-ionic surfactant [31]. Briefly, equal volume (1000 L) of gold nano-particles and Tween 20 (1.83 mg/mL, before mixing) in 10 mM sodium phosphate buffer (pH 7) were gently mixed and allowed to stand for a minimum of 30 min. Then, 1000 L of 0.5 mM 16-MHDA solution was added and the final mixture was kept undisturbed for 3 h for 16-MHDA to be chemisorbed onto the gold nano-particles. Unreacted excess thiols and Tween 20 were removed from alkanethiol modified gold nano-particles
188
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
by repeated centrifugation for 10 min at 3000 rpm followed by decantation of supernatants. The nano-particles were then resuspended in buffer. To prevent light-induced flocculation of the nano-particles and oxidation of alkanethiolates; all gold nanoparticles solutions were stored in the dark and refrigerated at 4 ◦ C [32]. 2.3.2. Coupling of O-phosphorylethanolamine on to gold nano-particles Carbodiimide chemistry is commonly used to form amide bonds between carboxylic acid and primary amines [33]. As a first step, EDC reacts with a carboxylic acid group and forms an amine reactive intermediate. In the next step, addition of a primary aminecontaining molecule in the reaction mixture leads to an amide linkage between amine and the activated carboxylic acid group. The activation is generally done at lower pH (pH ∼ 4.5). Since no aggregation of particles was observed at this pH, for all experiments a working pH of 4.5 was selected [34]. A 10−2 M stock solution of EDC was prepared in deionized water and stored below 0 ◦ C. The stock solution was used immediately to prevent extensive loss of activity. 10−4 M stock solution of O-phosphorylethanolamine (PEA) was prepared in deionized Millipore water and stored at 4 ◦ C. Both these solutions were equilibrated to room temperature before use. To 1 mL of MHDA modified-gold nano-particles, 100 L each of stock solution of PEA and EDC were added step wise under vigorous stirring. Stirring was continued for 20–30 s and then the nano-particles were stored undisturbed for 24 h. This step yielded nano-particles containing PEA, the specific ligand for CRP. Unreacted reagents were removed by repeated centrifugation and the collected nano-particles were resuspended in buffer as described above. 2.3.3. Interaction of CRP with functionalized gold nano-particles A stock solution of CRP was prepared by dissolving 0.18 mg CRP in 50 mL of 0.1 M calcium chloride solution. To 1 mL of functionalized gold nano-particles conjugated with PEA, 20 L of stock solution of CRP was added. All the measurements were made in triplicate and the average was taken. The relative optical density (OD) was corrected for dilution. The OD of the peak was measured by adding varying amounts of CRP. 2.3.4. Interaction of albumin with functionalized gold nano-particles Stock solution of albumin was prepared by dissolving 100 mg albumin in 100 mL of 10 mM phosphate buffer. To 1 mL of functionalized gold nano-particles, 10 L of stock solution of human albumin solution was added. To the same solution varied quantity of CRP was added to know the binding of CRP in the presence of albumin. In each case, the absorption spectra were recorded. 2.3.5. Estimation of CRP in blood serum using functionalized nano-particles Blood samples were collected from patients reported to Cardiology Department. The samples were collected according to the guidelines [35]. 1 mL of PEA conjugated gold nano-particles (PEAGNP) was taken and 10 L serum separated from the blood samples was added to the solution. The solution was shaken well for 10 s, and the plasmon peak was monitored. All the measurements were made in triplicate. 3. Results and discussion 3.1. Synthesis and surface modifications of gold nano-particles GNPs synthesised by the method of Turkevich et al. [30] yielded spherical particles with an average diameter of 39 ± 3 nm by the
Fig. 1. Optical absorption spectra of gold nano-particles and modified nanoparticles.
particle size analysis. Changes in the plasmon resonance peak monitored during each stage of modifications are shown in Fig. 1. The surface plasmon resonance peak of unmodified gold nano-particles is centered on 520 nm as reported earlier indicating that the nanoparticles were not aggregated but well dispersed as individual particles. GNPs modified with 16-MHDA exhibit a peak around 524 nm. The red shift in the position of the plasmon absorption is produced by a perturbation in the dielectric constant around the nano-particles due to chemisorption of 16-MHDA forming a selfassembled monolayer [31]. No significant broadening was observed after this step, indicating that the particles were not aggregated upon chemisorption of 16-MHDA. Further modification with EDC and PEA resulted in red shift to 526 nm. The results were further confirmed by TEM imaging, particle size analysis, potential measurement and FT-IR analysis. The generation of spherical nano-particles with nearly identical sizes was substantiated from the TEM image shown in Fig. 2A. Detectable variation in particle size was resulted by each modification steps as reflected in TEM images shown in Fig. 2B and C. Particle sizes and potential parameters summarized in Table 1 further support the results emerged from TEM. Particle sizes were steadily increased after each modification steps namely formation of 16MHDA layer followed by the coupling of PEA. Furthermore, the potential values (Table 1) altered after each steps indicating the modification of the particles. The net negative charge of the gold nano-particles reduced as result of the attachment of 16-MHDA. The size also increased further reflecting the presence of MHDA on the particle surface. Even after the modification, presence of net negative charge on the nano-particle surface stabilized them against aggregation in water. The net negative charge is slightly increased when PEA is attached onto the particles through the –COOH groups of 16-MHDA. In weakly acidic conditions the phosphate moiety in PEA exists in anionic form (PO4 − ). The potential value of O-phosphorylethanolamine at pH 4.5 was found to be −0.367 (Fig. 1S, Supplementary information). The increase in net negative charge of GNP modified with PEA could be assigned to the additional charge acquired from phosphate groups. To get further insight into the surface modifications of gold nano-particles, FT-IR spectroscopy was used. Nano-particles functionalized with 16-MDHA showed a strong band at 1734 cm−1 indicating the presence of carboxylic acid functionalities on the parTable 1 potential and hydrodynamic size of gold nano-particles and modified nanoparticles in solution as a function of surface modifications. System
potential (mV)
Effective diameter (nm)
As prepared gold nano-particles (citrate coated) 16-MHDA functionalized
−65 ± 2
39 ± 3
−17.6 ± 1
45 ± 1
Nano-particle PEA-GNP CRP–PEA-GNP
−23.5 ± 2 −4.65 ± 2
51 ± 1 107 ± 5
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
189
Fig. 2. TEM micrograph of gold nano-particles during each stages of modification. (A) Gold nano-particles; (B) after modification with 16-MHDA; (C) PEA-GNP.
ticles (Fig. 2S, Supplementary information). –CO-stretching band of 16-MDHA has nearly disappeared in the next step, namely the conjugation of PEA through the activated –COOH groups. The coupling was further confirmed by the peak centered on 1572 cm−1 reflecting the formation of –OCNH entity (Supplementary information).
3.2. Binding of CRP onto the modified nano-particles We studied the interaction of CRP with the functionalized nano-particles. The PEA-GNPs were equilibrated with solutions containing varied concentrations of CRP. The binding of CRP
Fig. 3. Colour changes of the nano-particles induced by surface modification. (A) Gold nano-particles, (B) gold nano-particles modified with 16-MHDA, (C) gold nano-particles attached to PEA and (D) after addition of CRP.
190
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
Fig. 4. Optical absorption spectra of PEA-GNP in the presence of varied concentrations of CRP.
resulted in a visibly observable colour change, reflecting shift in the plasmon resonance absorption (Fig. 3). The optical absorption spectra of the nano-particles in the presence of varied amounts of CRP are shown in Fig. 4. A red shift of the plasmon peak with concomitant broadening of the spectrum occurred in a concentration-dependent manner. The shift and broadening attained maximum values at intermediate concentrations of CRP. CRP causes a change in the absorbance since it induces aggregation of gold nano-particles through specific molecular interactions with PEA. The degree of aggregation is concentration dependent with maximum aggregation occurring for intermediate concentration of CRP in the system. If only a small number of CRP molecules are present in the system, aggregation will not proceed to a significant extent because of the limited availability of the molecule that bridges nano-particle together. Only at intermediate concentration of CRP, when one CRP molecule can bind several gold nano-particles aggregation occurs since a single CRP molecule has five binding sites. When the concentration of CRP is large enough, PEA groups become unavailable for CRPmediated aggregation. Beyond a certain amount of CRP, the solution turned turbid indicating clumping of particles. Earlier studies have shown similar results, namely the aggregation of nano-particles mediated by proteins [36]. We used particle size measurement and TEM imaging to get further insight in to the process. The effective diameter of the protein bound particles increased to 107 ± 5 and zeta potential is reduced to −4.65 ± 2 (see Table 1). The reduced charge together with the increased size facilitates interaction of the particles resulting in the aggregation. TEM micrographs of PEAGNP after the addition of lower and higher concentrations of CRP are shown in Fig. 5A and B. The number of particles in the cluster increases significantly, when PEA-GNP was allowed to interact with CRP. The formation of bigger clusters was a manifestation of specific molecular recognition occurring between the multivalent CRP and PEA group in PEA-GNP. CRP molecules can link several PEA-GNP nano-particles together to form larger clusters. The aggregation with increase in CRP concentration was reasoned due to the bridging of nano-particles by the protein. We have drawn such a conclusion by visualizing the particles depicted in TEM (Fig. 5B). We recorded the infrared spectrum of the particles after equilibrating with CRP. The IR spectrum showed additional peaks at 1650 cm−1 and 1540 cm−1 , characteristic of amide I and amide II bands in protein chains, strongly suggesting the adsorption of CRP onto the PEA-GNP (Fig. 3S, Supplementary information). 3.3. Quantification of CRP The optical density measurements of PEA-GNPs at different CRP concentrations were monitored. It was found that with increase in CRP concentration the plasmon resonance peak showed red shift with a concomitant decrease in OD. Red shift in absorption peak and a reduction in intensity with concentration of analytes are well documented [37]. The change in OD values was increased initially and
Fig. 5. TEM micrograph of gold nano-particles after interacting with CRP. (A) On addition of lower concentration of CRP; (B) on addition of higher concentrations of CRP.
then levelled off indicating the saturation of available binding sites for CRP (Fig. 4S, Supplementary information). A linear relationship between OD of the plasmon peak and concentration was observed until CRP was bound onto the entire available PEA. Further addition of CRP did not show any variation in OD (Fig. 4S, Supplementary information). We observed that GNPs were started aggregating (slight turbidity) when the concentration of CRP exceeded 500 ng mL−1 and the solution became completely turbid beyond 600 ng mL−1 of CRP. Larger concentrations of CRP (>600 ng mL−1 ) caused the aggregation of almost all PEA-GNPs indicating the saturation of available binding sites for CRP. The change in absorption reaches maximum with the addition of 450–500 ng mL−1 of CRP. This observation led us to conclude that, this methodology can be used for the estimation of CRP in solutions having a concentration of ≤450 ng mL−1 . It was observed that PEA-GNPs are separated from each other before the addition of CRP, and they become closer and closer with the addition of increased amount of CRP leading to the clustering of particles. For quantification of CRP, we constructed a calibration plot using the concentrations and corresponding OD (Fig. 5S, Supplementary information).
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
191
Table 2 Quantitative data of CRP in serum—comparison of the methods. Sample code
Amount of CRP (nephelometry)a (mg L−1 )
Amount of CRP using (PEA-GNP) (mg L−1 )
A B C D E F G
39.07 22.72 16.79 12.76 11.45 Negative Sample Negative Sample
38.97 ± 0.21 22.45 ± 0.10 16.05 ± 0.50 12.15 ± 0.09 11.32 ± 0.12 No measurable shift in the peak No measurable shift in the peak
a
Immunoassay (average of three measurements).
3.4. Interaction of albumin with the modified particles Albumin is one of the major blood proteins. Non-specific adsorption of albumin onto the particles is known to be a major problem associated with the usage of functionalized gold nanoparticles. Albumin concentration is several folds higher than the concentration of CRP in the blood. Interference of albumin in the measurement, if any, has to be assessed. To the solution of PEA-GNPs, 10 L of stock solution of albumin was added and the plasmon resonance absorption was monitored. The plasmon absorption peak of PEA-GNP was centered at 526 nm. When albumin was added to the above solution, the plasmon peak is again centered at 526 nm indicating that albumin was not binding on to the ligand (Fig. 6S, Supplementary information). Park et al. have recently shown that binding of albumin onto polymeric nano-particles could be reduced remarkably by modifying the surface by phosphorylcholine moieties [17]. Also the isoelectric point of albumin is ∼4–4.5. At a pH below their isoelectric point, proteins carry a net positive charge; above their isoelectric point they carry a net negative charge. Since the working pH was 4.5 albumin is supposed to be negatively charged and hence adsorption of albumin onto the negatively charged PEA-GNP is unlikely. Interestingly, when CRP was added to the same solution, plasmon resonance peak shifted to 529 nm strongly suggesting the binding of CRP to its ligand even though albumin was present (Fig. 7S, Supplementary information). This observation further confirm that albumin is neither affecting the plasmon absorption peak nor interfering the interaction of CRP with its ligand. The plasmon absorption peak showed red shifts with increase in the concentration of CRP when additional amount of CRP was added into the same solution (Supplementary information). 3.5. Measurement of CRP in real samples Blood samples collected from various patients were tested using PEA-GNP and the results were compared with those obtained by normal CRP diagnostic test using CRP specific antibody. The clinical assay for CRP detection is generally carried out by immunoassay using nephelometry with detection limit of ∼3.5 g mL−1 . Serum samples containing <3.5 g mL−1 of CRP, as per this method, is considered as negative. Higher concentration of CRP (>3.5 g mL−1 ) is treated as higher risk factor for myocardial infraction and those samples are designated as positive samples. Here we provide a comparative study of CRP estimation using nephelometry method and PEA-GNPs. PEA-GNPs showed plasmon absorption peak around 526 nm (Fig. 6A). The samples designated as negative CRP samples did not show any considerable shift in the plasmon peak indicating very low concentration of CRP (Fig. 6A, Sample F in Table 2, see the overlapped traces). The plasmon peak, however, shifted to higher wavelength when serum termed as positive is added into the functionalized nano-particles reflecting the sensing of CRP by the particles. The plasmon peak was shifted to 529 nm for CRP concentration in the range of 11–12 g mL−1 (spectra D and E in Fig. 6B and Samples D and E in Table 2). The absorption
spectra of these two samples are almost overlapped since the CRP content is very close. With serum containing higher concentrations of CRP the spectra showed further shift with a decrease in intensity (spectra A, B and C, Fig. 6B and Samples A, B and C in Table 2). These samples were diluted several times in order to bring them in the sensing range. Table 2 compares quantitative data generated using both methods. The results emerged from the current method compare well with immunoassay method, indicating the reliability of the newly proposed approach using functionalized nano-particles. One of the remarkable features of the present method is its sensitivity comparing to the method based on nephelometry. The present approach enables the measurement in the nanogram range while the immunoassay is in the microgram range. The volume of serum used in the measurement was 10 L. Many serum samples containing higher quantity of CRP (as per immunoassay) were diluted prior to the analysis using the current approach since higher quantity of CRP led to the aggregation of PEA-GNPs. Several recent studies confirm that elevated serum level of plasma CRP is associated with an increased risk of experiencing myocardial infraction and sudden cardiac death in apparently healthy subjects. These studies undoubtedly suggest that elevated CRP concentration may predict a higher risk for future cardiovascular diseases [1–3]. Very recently Zhan and Bard reported a chemiluminescence method for the determination of CRP using polyclonal human CRP antibodies incorporated in liposomes [38]. These approaches though sensitive require antibodies as recognition element for CRP. The effort in this communication was to develop a methodology avoiding antibodies. Towards this goal we synthesised gold nanoparticles and functionalized the surface of gold nano-particles with CRP specific ligand. The data emerged from the study suggest that appropriately functionalized gold nano-particles can be employed
Fig. 6. (A) Optical absorption spectra of PEA-GNP before and after the addition of 10 L negative serum. (B) Optical absorption spectra of PEA-GNP (traces A, B, C, D and E) after adding 10 L serum samples.
192
V. Raj, K. Sreenivasan / Analytica Chimica Acta 662 (2010) 186–192
for the assay of CRP in serum. A visibly observable colour change with positive samples is another added advantage of the present method in the sense that it could be used for the quick screening of positive and negative samples. 4. Conclusion We have developed a new colorimetric sensing approach employing functionalized gold nano-particles for the estimation of CRP in serum with a high degree of selectivity and sensitivity. Gold nano-particles were synthesised and functionalized with PEA, the specific ligand of CRP. The nano-particles were characterized using TEM, particle size, potential measurements, FT-IR and optical absorption spectroscopy. The plasmon absorption spectra of PEA-GNPs were found to show a considerable red shift in the presence of CRP reflecting its binding onto the particles. potential and particle sizes were found to vary in accordance with the modification and CRP binding. The concentration depended shift facilitated the measurement of CRP in serum. The binding of CRP onto the nano-particles was clearly visible with a distinct colour change. The newly designed approach was used to estimate the CRP in serum of patients and the values were compared well with those generated using immunoassay technique. The method involves simple synthetic approaches using commercially available components and is devoid of antibodies. Acknowledgements We wish to thank Dr. C.P. Sharma and Willy Paul for zeta potential and particle size measurement, Dr. Annie John for TEM analysis and Ms. Moly Antony for providing blood samples. Vidya Raj thanks Council of Scientific and Industrial Research (CSIR), India for the senior research fellowship.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
[21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]
Appendix A. Supplementary data
[33] [34] [35]
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.aca.2010.01.007.
[36] [37] [38]
P.M. Ridker, J.E. Buring, N.R. Cook, N. Rifai, Circulation 107 (2003) 391. W.T. Tillet, W.T.J. Francis, J. Exp. Med. 52 (1930) 561–567. I. Kushner, M.L. Broder, D.J. Karp, J. Clin. Invest. 61 (1978) 232–235. P.M. Ridker, Circulation 107 (2003) 363–369. Q. Zen, W. Zhong, R.F. Mortensen, J. Cell. Biochem. 64 (1997) 140–151. J.E. Volanaski, Mol. Immunol. 38 (2001) 189–197. J. Whicher, L. Biasucci, N. Rifai, Clin. Chem. Lab. Med. 37 (1999) 495–503. O. Gul, E. Calay, U. Sezerman, H. Basaga, Y. Gurbuz, Sens. Actuators B 126 (2007) 581–587. M.H.F. Meyer, M. Hartman, Biosens. Bioelectron. 21 (2006) 1987–1992. C. Glad, A.O. Grubb, Anal. Biochem. 85 (1978) 180–187. K. Kriz, F. Ibraimi, M. Lu, L. Hansson, D. Kriz, Anal. Chem. 77 (2005) 5920–5924. W. Miao, A.J. Bard, Anal. Chem. 75 (2003) 5825–5834. H.Y. Tsai, C.F. Hsu, I.W. Chiu, C. Bor Fuh, Anal. Chem. 79 (2007) 8416–8419. R.E. Banks, Crit. Rev. Clin. Lab. Sci. 37 (2000) 131–182. J.L. Bock, Am. J. Clin. Pathol. 113 (2000) 628–646. P.M. Ridker, M. Cushman, M.J. Stampfer, R.P. Tracy, C.H. Hennekens, Circulation 97 (1998) 425–428. A. Park, S. Kurosawa, J. Watanabe, K. Ishihara, Anal. Chem. 76 (2004) 2649–2655. C.M. Merritt, J.W. Winkelman, Anal. Chem. 61 (1989) 2362–2365. V. Raj, P.R. Hari, K. Sreenivasan, Anal. Chim. Acta 592 (2007) 45–50. V.H. Perez-Luna, K. Aslan, P. Betala, Colloidal gold, in: J.A. Schwarz, C.I. Contescu, K. Putyera (Eds.), Encyclopedia of Nanoscience and Nanotechnology, vol. 2, Marcel Dekker, New York, USA, 2004, pp. 27–49. U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters. Springer Series in Material Science 25, Springer, New York, USA, 1995. M.A. Hayat (Ed.), Colloidal Gold: Principles, Methods, and Applications, Academic Press, San Diego, CA, USA, 1989. M.C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293–346. E. Katz, I. Willner, Angew. Chem. Int. Ed. 43 (2004) 6042–6108. W. Knoll, Annu. Rev. Phys. Chem. 49 (1998) 569–638. N. Nath, A. Chilkoti, Anal. Chem. 74 (2002) 504–509. N. Nath, A. Chilkoti, Anal. Chem. 76 (2004) 5370–5378. J.J. Li, Y.A. Wang, W. Guo, J.C. Keay, T.D. Mishima, M.B. Johnson, X. Peng, J. Am. Chem. Soc. 125 (2003) 12567–12611. E. Nandan, N.R. Jana, J.Y. Ying, Adv. Mater. 20 (2008) 2068–2073. J. Turkevich, P.C. Stevenson, J. Hillier, Discuss. Faraday Soc. 11 (1981) 55–75. K. Asian, V.H. Perez-Luna, Langmuir 18 (2002) 6059–6065. C.S. Weisbecker, M.V. Merritt, G.M. Whitesides, Langmuir 12 (1996) 3763–3772. J. Lee, A.O. Govorov, J. Dulka, N.A. Kotov, Nano Lett. 4 (2004) 2323–2327. A. Gole, C.J. Murphy, Langmuir 21 (2005) 10756–10762. S.M. Lewis, A. Koepke, Haematology Laboratory Management and Practice, American Chemical Society, Washington, DC, 1995. K. Asian, C.C. Luhrs, V.H. Perez-Luna, J. Phys. Chem. B 108 (2004) 15631–15639. K. Ai, Y. Liu, L. Lu, J. Am. Chem. Soc. 131 (2009) 9496–9497. W. Zhan, A.J. Bard, Anal. Chem. 79 (2007) 459–463.